Does Narcolepsy Symptom Severity Vary According to HLA-DQB1*0602 Allele Status?

Abstract

Study Objectives:

To investigate associations between HLA-DQB1*0602 allele status and measures of narcolepsy symptom severity.

Design:

Cross-sectional study of population-based narcolepsy patients.

Setting:

King County, Washington.

Participants:

All prevalent cases (n = 279) of physician-diagnosed narcolepsy ascertained from 2001-2005.

Interventions:

N/A

Measurements:

Narcolepsy diagnosis was based on cataplexy status, diagnostic sleep study results, and chart review. The number of HLA-DQB1 alleles was determined from buccal genomic DNA. Symptom severity instruments included the Epworth Sleepiness Scale (ESS), the Ullanlinna Narcolepsy Scale (UNS), age of symptom onset, subjective sleep latency and duration, and various clinical sleep parameters. We used linear regression adjusted for African American race and an extended chi-square test of trend to assess relationships across ordered groups defined by allele number (0, 1, or 2).

Results:

Narcolepsy patients were 63% female and 82% Caucasian, with a mean age of 47.6 years (SD = 17.1). One hundred forty-one (51%) patients had no DQB1*0602 alleles; 117 (42%) had one; and 21 (7%) had two. In the complete narcolepsy sample after adjustment for African American race, we observed a linear relationship between HLA-DQB1*0602 frequency and sleepiness as defined by the ESS (P < 0.01), narcolepsy severity as defined by UNS (P < 0.001), age of symptom onset (P < 0.05), and sleep latency (P < 0.001). In univariate analyses, HLA-DQB1*0602 frequency was also associated with napping (P < 0.05) and increased car and work accidents or near accidents (both P < 0.01). Habitual sleep duration was not associated with HLA status. These race-adjusted associations remained for the ESS (P < 0.05), UNS (P < 0.01), and sleep latency (P < 0.001) when restricting to narcolepsy with cataplexy.

Conclusions:

Narcolepsy symptom severity varies in a linear manner according to HLA-DQB1*0602 allele status. These findings support the notion that HLA-DQ is a disease-modifying gene.

Citation:

Watson NF; Ton TGN; Koepsell TD; Gersuk VH; Longstreth WT. Does narcolepsy symptom severity vary according to HLA-DQB1*0602 allele status? SLEEP 2010;33(1):29-35.

NARCOLEPSY IS A NON-PROGRESSIVE SLEEP DISORDER WITH US BASED PREVALENCE ESTIMATES BETWEEN 3 AND 67 PATIENTS PER 100,000.^1–5 HALLMARKS include sleepiness and REM sleep phenomena that intrude into wakefulness such as cataplexy, hypnagogic/ hypnopompic hallucinations, and sleep paralysis.⁶ Nocturnal sleep is often disturbed, and precipitous daytime napping is common. The disorder often strikes in adolescence or early adulthood affecting all aspects of life from family and interpersonal relationships to school and workplace performance.⁷ The etiology of narcolepsy is unknown, but observations of hypocretin producing cell loss in the dorsolateral hypothalamus,^8,9 reduced cerebrospinal fluid hypocretin levels,¹⁰ season of birth associations,^11,12 a strong HLA association,^13,14 and a recent genome-wide association study revealing associated polymorphisms in the T-cell receptor α locus point to an autoimmune mechanism.¹⁵

Class II HLA molecules such as DQB1*0602 initiate the adaptive immune response by presenting short, pathogen-derived peptides to T cells. At times, self-peptides are presented generating self-activated T cells causing an autoimmune response, observed in diseases such as multiple sclerosis (MS),^16,17 rheumatoid arthritis,¹⁸ type I diabetes mellitus,¹⁹ Alzheimer disease,²⁰ and Behçet disease.²¹ Transgenic animal models provide direct evidence implicating HLA molecules in autoimmune disease.^22,23 Microsatellite and HLA haplotype analysis suggest that HLA-DQB1*0602 has direct involvement in susceptibility to narcolepsy, particularly those with cataplexy.^12,24,25 Nine HLA class II alleles carried in trans with DQB1*0602 influence disease predisposition,¹⁴ and the relative risk of developing narcolepsy is 2–4 times as high in HLA-DQB1*0602 homozygotes as in heterozygotes.²⁴ The HLA complex can also modify disease expression. For example, the HLA-DR2 allele is associated with earlier age of onset and clinical disability ratings in MS, with heterozygosity predisposing the more severe phenotype.^26,27 This allele also modulates relapsing-remitting MS by precipitating the secondary progressive phase, and is associated with disease severity as defined by neuroimaging and neurocognitive measures.^16,28 Similarly, the HLA-B genotype modulates the clinical course of Behçet disease including progression of ocular lesions, articular symptoms, skin lesions, and neurological symptoms²¹; and the HLA-E genotype influences age of onset of type I diabetes mellitus.¹⁹

Despite the strong association and dose effect of the HLA-DQB1*0602 genotype on narcolepsy susceptibility, researchers have been unable to consistently demonstrate disease-modifying characteristics for this allele in patients with narcolepsy.^12,24 The goal, therefore, of this study is to investigate the association between HLA-DQB1*0602 allele status and measures of narcolepsy disease severity including the Epworth Sleepiness Scale (ESS), the Ullanlinna Narcolepsy Scale (UNS), age of symptom onset, subjective sleep latency and duration, and various clinical sleep parameters in a population-based sample of individuals with narcolepsy. Associations between HLA-DQB1*0602 and narcolepsy disease severity would support the notion that HLA-DQ is a disease-modifying gene.

METHODS

Narcolepsy Case Ascertainment

Narcolepsy patients were obtained from a parent study investigating prevalence and incidence rates in King County, Washington.⁵ Patients were identified using multiple overlapping methods, some of which were directed at clinicians and some at patients. We met with directors of sleep disorders centers to explain the study and seek their help. During the course of parent study recruitment (July 2001 to June 2005) we sent monthly letters to clinicians working in sleep disorder centers and to all neurologists in King County asking them to inform us of any patient whom they diagnosed with narcolepsy. We also had a single mailing to the County's family medicine physicians and to psychiatrists. Finally we had a mailing to the community clinics in King County where patients without financial resources often receive their care.

With respect to direct appeals to patients, study fliers were posted in waiting rooms of sleep disorder centers. Presentations about the parent study were made at the region's support groups and at the Narcolepsy Network's national annual meeting held in King County in 2004. During recruitment, a monthly newsletter was sent to all who had participated in the parent study. Spreading news about the study by word of mouth was encouraged. Pharmacists in King County agreed to include an information sheet about the parent study with all prescriptions for certain medications commonly used to treat patients with narcolepsy. Advertisements and public service announcements were placed in several newspapers, on County buses, on several radio stations, and on television. All of these sources of information included a contact telephone number and address for the parent study's web site. The University of Washington also maintained a web site for research volunteers where information about the study was available. The Institutional Review Board at the University of Washington approved the study.

Patients with physician-diagnosed narcolepsy identified by any of these means were entered into a registry and invited to participate in the parent study. Patients identified by clinicians were invited through their clinician, while patients contacting study personnel were invited directly. Participating patients provided written informed consent and were interviewed by a trained professional who inquired about clinical manifestations, onset of disease, and diagnosis. Medical records were requested and abstracted by the study neurologist (WL). Information regarding cataplexy came from interviews, including screening questions from a well-validated cataplexy algorithm²⁹ and review of medical records. In screening for cataplexy, patients were asked, “Do you currently experience, or have you ever experienced, episodes of muscle weakness, such as weakness in your legs or buckling of your knees, during the following situations: (a) When you tell or hear a joke? (b) When you laugh?” Cataplexy was classified as present when the respondent answered yes to (a), or no to (a) but yes to (b).²⁹ For this study, we defined cataplexy as present if indicated by positive screen or medical record review. Narcolepsy with cataplexy was defined according to the criteria of the International Classification of Sleep Disorders (ICSD-2).³⁰ The methods described above identified 282 patients for interview, 279 of whom provided written informed consent, completed the interview, and provided buccal specimens for HLA typing. Review of medical records indicated that all 279 patients (100%) had experienced excessive daytime sleepiness, 158 (57%) had experienced cataplexy, and 138 (49%) carried at least one HLA-DQB1*0602 allele. Eighty-five percent of patients were either diagnosed by a sleep medicine specialist, and/or were cataplexy positive, and/or underwent formal diagnostic sleep studies. Diagnostic testing status was either unknown or absent in the remaining 15% of physician-diagnosed narcolepsy patients.

Symptom Severity Rating Instruments

Trained professionals administered in-person interviews to narcolepsy patients. Subjects were asked, on average, “About how many hours would you estimate you actually spend asleep each night,”³¹ and “About how many minutes would you estimate there are from when you go to bed until when you go to sleep each night?” Sleepiness and overall narcolepsy symptom severity were ascertained with 2 well-known validated instruments, the 8-item ESS³² and 11-item UNS.³³ In both cases, higher scores indicated greater sleepiness and worse narcolepsy symptom severity, with an ESS score maximum of 24, and a UNS maximum of 44. Age of symptom onset was revealed by the question, “In what year did you first experience any symptom of narcolepsy.” Symptom onset was calculated from the patient's birth date. In 10 patients, symptom onset was either established or confirmed via medical chart review by the study neurologist (WL).

Additional clinical symptom severity was ascertained by asking subjects, “In the past year, how often, on average: 1) do you have trouble falling asleep; 2) do you wake up during the night and have trouble getting back to sleep; and 3) get so sleepy during the day or evening that you have to take a nap?”³⁴ Subjects were then asked: 1) “do you have dream-like experiences when you are falling asleep or waking up, even though it seems like you are awake; 2) do you ever find yourself paralyzed or unable to move just as you are falling asleep or waking up; 3) do you have uncontrollable or irresistible urges to fall asleep or nap during the day, or find yourself accidentally falling asleep when you do not want to; 4) have you ever had an accident or near accident due to drowsiness or sleepiness while driving; and 5) have you ever had an accident or near accident due to drowsiness or sleepiness while at work?” Response categories included never, seldom, occasionally, often, or always.³⁵

HLA-DQB1*0602 Allele Typing

Genotyping of HLA-DQB1 entailed quantitative DNA amplification and fluorescence detection with sequence-specific probes, as described previously for HLA-DRB1 alleles.³⁶ Briefly, primers and probes were designed within exon 2 of the HLA-DQB1 locus based on the sequence alignments of the IMGT/HLA Sequence Database (http://www.ebi.ac.uk/imgt/hla/align.html) to distinguish major DQB polymorphisms at codons 25-28 and 47-48, which identify allele DQB1*0602 and distinguish several additional DQB1 alleles. Primers and probes for HLA-DRA were used to control for template quality and quantity, as previously described.³⁶ Approximately 30 nanograms of genomic DNA in TaqMan Universal Master Mix (Applied Biosystems, Inc.) in a final volume of 25 microliters per well were combined with primers and probes and amplified on an ABI PRISM® 7000 Sequence Detection System (Applied Biosystems, Inc.) for 45 cycles.³⁷ Presence or absence and number of DQB1*0602 alleles (0, 1, or 2) was determined as previously described.³⁷

Statistical Analysis

To test our hypothesis that narcolepsy symptom severity and clinical characteristics are associated with HLA DQB1*0602 allele number, we conducted multiple linear regression analysis with robust standard error estimates. In all analyses, HLA-DQB1*0602 status was the ordinal independent variable. We analyzed the total narcolepsy sample, as well as a subsample restricted to cataplexy positive patients. We present results as African American race-adjusted mean values and 95% confidence intervals. Ordinal dependent variables with response categories never, seldom, occasionally, often, or always were collapsed into 3 groups: never, occasional (seldom and occasionally), and frequent (often and always). For dichotomous dependent variables we used a χ² test for trend. For continuous demographic variables we used analysis of variance (ANOVA) planned contrasts. For dependent variables with ≥ 3 ordered categories, we calculated χ² test statistics with one degree of freedom based on a modified χ² test for trend for larger contingency tables.³⁸ We used Stata 10 statistical package for all analyses (StataCorp LP, College Station, TX, USA).

RESULTS

The narcolepsy sample (n = 279) was 63% female and aged 47.6 years (SD = 17.1). Race was 82% Caucasian, 9% African American, and 9% other. Seventy-eight percent had a college education or higher. With the exception of African American race, demographics were similar when stratified by HLA-DQB1*0602 allele status. Sleepiness was currently being treated in the majority of subjects with stimulants or sodium oxybate regardless of HLA status, and cataplexy and HLA haplotype were tightly linked (Table 1).

Table 1.

Demographic characteristics of 279 physician-diagnosed narcolepsy patients from King County, Washington

	Number of HLA-DQB1*0602 Alleles
	None (n = 141)		One (n = 117)		Two (n = 21)		P-value*
Age, mean, (SD)	45.3	(15.8)	50.2	(18.0)	48.7	(19.3)	0.40
Male sex, n (%)	90	(63.8)	72	(61.5)	14	(66.7)	0.95
African American race, n (%)	8	(5.7)	12	(10.3)	6	(28.6)	<0.01
Cataplexy, n (%)	58	(41.1)	85	(72.6)	15	(71.4)	<0.0001
Current stimulant use†, n (%)	129	(91.5)	114	(97.4)	20	(95.2)	0.09
Annual income, n (%)
<$20,000	28	(21.9)	26	(24.1)	6	(33.3)	0.15
$20,000–$39,000	28	(21.9)	28	(25.9)	4	(22.2)
$40,000–$59,000	18	(14.1)	20	(18.5)	2	(11.1)
≥ $60,000	54	(42.2)	34	(31.5)	6	(33.3)
Education, n (%)
High school or technical school	27	(19.4)	32	(27.4)	3	(14.3)	0.12
College	78	(56.1)	65	(55.6)	16	(76.2)
Graduate school	34	(24.5)	20	(17.1)	2	(9.5)

^*tests used were ANOVA planned contrast for continuous variables, χ² test of trend for dichotomous variables, and a modified χ² test of trend for larger ordinal contingency tables;

^†Current stimulant use defined as use of one of the following: amphetamine, dextroamphetamine, methamphetamine, methylphenidate, pemoline, modafinil, or sodium oxybate.

In the total narcolepsy sample, including patients with and without cataplexy, we found a significant linear trend between HLA-DQB1*0602 status and the ESS (P < 0.01), UNS (P < 0.001), age of symptom onset (P < 0.05), and subjective sleep latency (P < 0.001) adjusted for African American race. Values for the ESS and UNS increased in a dose response manner with allele number, whereas age of symptom onset and sleep latency exhibited an inverse dose response relationship with increasing number of alleles (Table 2). In the cataplexy-restricted subsample, the race-adjusted HLA associations persisted for the ESS (P < 0.05), UNS (P < 0.01), and subjective sleep latency (P < 0.001), but not for age of symptom onset (P = 0.84). Again, the number of alleles was associated with an increasing trend in ESS and UNS means, with an inverse dose response association with sleep latency (Table 2). Subjective sleep duration was not associated with HLA status in either the complete or cataplexy-restricted sample. In a univariate analysis with other narcolepsy symptoms and clinical factors, we found a significant dose response relationship between the number of HLA-DQB1*0602 alleles and the frequency of subjective napping (P < 0.05), with 80% of patients with 2 alleles reporting frequent napping, compared to 62% with no alleles. HLA allele number was also positively associated with the frequency of workplace and roadway accidents or near accidents, with 76% of subjects with 2 HLA-DQB1*0602 alleles reporting occasional or frequent accidents or near accidents due to drowsiness or sleepiness while driving (P < 0.01) and 48% reporting accidents or near accidents while at work (P < 0.01; Table 3).

Table 2.

Narcolepsy symptom severity by HLA-DQB1*0602 allele status in all narcolepsy patients and a cataplexy positive subset

	All narcolepsy patients (n = 279)*				Narcolepsy with cataplexy (n = 158)
	Number of HLA-DQB1*0602 alleles (mean, 95% CI)				Number of HLA-DQB1*0602 alleles (mean, 95% CI)
	No copies (n=141)	One copy (n=117)	Two copies (n = 21)	P-value‡	No copies (n = 58)	One copy (n = 85)	Two copies (n = 15)	P-value‡
Epworth Sleepiness scale	14.8 (14.0, 15.7)	16.2 (15.3, 17.0)	18.0 (16.6, 19.4)	<0.01	15.8 (14.6, 17.0)	17.1 (16.2, 18.0)	18.1 (16.3, 19.9)	<0.05
Ullanlina Narcolepsy Scale	11.7 (10.5, 12.8)	15.8 (14.4, 17.2)	17.5 (13.8, 21.3)	<0.001	14.8 (13.0, 16.5)	18.2 (16.8, 19.6)	20.9 (16.8, 24.9)	<0.01
Age of symptom onset (years)	19.2 (17.1, 21.3)	17.3 (15.2, 19.4)	13.3 (10.8, 15.7)	<0.05	15.1 (13.0, 17.2)	15.6 (13.6, 17.7)	13.9 (11.0, 16.8)	0.84
Sleep latency (minutes)	13.0 (10.3, 15.8)	8.5 (6.5, 10.5)	3.7 (2.2, 5.2)	<0.001	15.6 (10.8, 20.4)	6.9 (5.1, 8.7)	4.0 (2.4, 5.7)	<0.001
Sleep duration (hours)	6.7 (6.4, 7.0)	6.6 (6.3, 7.0)	6.5 (5.5, 7.5)	0.64	6.6 (6.1, 7.2)	6.6 (6.3, 7.0)	6.5 (5.2, 7.9)	0.92

^*three missing data from the Epworth, Ullanlinna, sleep latency, and sleep duration.

^‡linear regression adjusted for African American race

Table 3.

Frequency of narcolepsy symptoms and associated factors stratified by HLA-DQB1*0602 allele status

Narcolepsy characteristics	Number of HLA-DQB1*0602 alleles			P-value for trend test†
	None	One	Two
Hypnagogic/hypnopompic hallucinations	(n = 138)	(n = 117)	(n = 21)	0.55
Never (%)	11.6	15.4	4.8
Occasional (%)	54.4	46.6	57.1
Frequent (%)	34.1	38.5	38.1
Sleep paralysis	(n = 138)	(n = 117)	(n = 21)	0.30
Never (%)	32.6	30.8	14.3
Occasional (%)	47.8	47.0	61.9
Frequent (%)	19.6	22.2	23.8
Sleep attacks	(n = 138)	(n = 117)	(n = 21)	0.36
Never (%)	5.8	9.4	4.8
Occasional (%)	39.9	31.6	28.6
Frequent (%)	54.4	59.0	66.7
Sleep onset insomnia	(n = 137)	(n = 117)	(n = 21)	0.24
Never (%)	23.4	31.6	28.6
Occasional (%)	50.4	45.3	52.4
Frequent (%)	26.3	23.1	19.1
Sleep maintenance insomnia	(n = 137)	(n = 117)	(n = 21)	0.25
Never (%)	26.3	28.2	19.1
Occasional (%)	51.8	44.4	47.6
Frequent (%)	21.9	27.4	33.3
Napping	(n = 138)	(n = 117)	(n = 20)	<0.05
Never (%)	4.4	1.7	5.0
Occasional (%)	33.3	23.1	15.0
Frequent (%)	62.3	75.2	80.0
Motor vehicle accident or near accident	(n = 137)	(n = 117)	(n = 21)	<0.01
Never (%)	38.0	34.2	23.8
Occasional (%)	58.4	58.1	52.4
Frequent (%)	3.7	7.7	23.8
Work accident or near accident	(n = 138)	(n = 117)	(n = 21)	<0.01
Never (%)	72.5	67.5	52.4
Occasional (%)	23.9	23.9	23.8
Frequent (%)	3.6	8.6	23.8

^†based on a modified χ² test for trend for larger ordinal contingency tables

DISCUSSION

We found that narcolepsy disease characteristics and measures of symptom severity varied in a dose response manner according to HLA-DQB1*0602 status. This was true for sleepiness as measured by nap propensity and the Epworth Sleepiness Scale, and abnormal sleeping tendency and cataplexy as measured by the Ullanlinna Narcolepsy Scale. In addition, age upon first narcolepsy symptom and subjective sleep latency varied according to allele status. To confirm our findings we limited our analysis to only those narcolepsy patients with concurrent cataplexy, thus ensuring patients fulfilled the ICSD-2 diagnostic criteria³⁰ regardless of the presence or absence of objective testing. Again, we found a dose response relationship between HLA-DQB1*0602 allele status and the Epworth and Ullanlinna Scales and sleep latency, suggesting that this HLA allele is not only associated with narcolepsy, but also modifies the disease phenotype.

Previous data assessing the influence of HLA status on sleep phenotypes have been mixed. Mignot et al. demonstrated that HLA-DQB1*0602 is associated with shortened REM latency, increased sleep efficiency, and decreased stage 1 sleep in normal subjects.³⁹ Similar to our results, Hong et al. revealed shorter sleep latencies, higher ESS scores, and more severe cataplexy symptoms in narcolepsy with cataplexy patients according to HLA-DQB1*0602 status.⁴⁰ However, an analysis of the influence of HLA status on these phenotypes in the total narcolepsy sample was not performed. Sasai et al. found increased ESS scores in predominantly HLA-DQB1*0602 positive narcolepsy with cataplexy patients when compared to HLA-negative narcolepsy without cataplexy patients but did not find a difference in ESS scores between HLA-positive and negative narcolepsy without cataplexy patients.⁴¹ Other studies investigating HLA status on narcolepsy symptom severity found no association with age of symptom onset, sleepiness, sleep latency, or cataplexy severity.^12,24 We found no association between allele status and habitual sleep duration, an issue previous studies had not considered.

Differences between these studies and ours are likely multifactorial in origin. First, we stratified HLA-DQB1*0602 status by allele number (0, 1, or 2) and analyzed the data in the overall narcolepsy sample as well as the narcolepsy with cataplexy subsample. Other studies dichotomized HLA status and performed comparisons in groups defined by cataplexy status.^12,24,40,41 Second, our study was population based, and therefore free from recruitment bias. Third, our study was based in the US, with likely differences in HLA-DQB1*0602 allele frequency between our population and populations from other countries with different racial makeups. Lastly, narcolepsy definitions were not uniform across studies.

Number of DQB1*0602 alleles was also associated with workplace and roadway safety, a finding with potential social implications. Unlike epilepsy, where commercial vehicle operation is prohibited and driving may be restricted following a seizure, federal or state regulations limiting the driving of patients with narcolepsy are nonexistent, despite the risks of precipitous sleep and cataplexy associated weakness while driving. Privacy issues and lack of validation currently obviate genotyping as a tool to inform workplace safety policy. Nevertheless, our results suggest that well-controlled prospective studies assessing workplace and roadway safety as a function of HLA status in patients with narcolepsy has potential public health implications and is worthy of consideration. The influence of HLA genotype on workplace safety also has implications for disability claims, insurance policy, and workers rights. These issues are fraught with numerous ethical, social, and legal implications, and more research is needed before HLA genotyping can be considered an accident risk assessment tool in narcolepsy patients, but as genomics becomes more commonplace in clinical and public policy arenas, genotyping of HLA or other disease-modifying genes may have a role to play in shaping public policy toward disorders such as narcolepsy in the future.

The dose effect of HLA-DQB1*0602 on narcolepsy symptom severity dovetails with the notion of narcolepsy as an autoimmune disease. Although firm establishment of narcolepsy as an autoimmune disease has been elusive,^13,42 a recent genome-wide association study (GWAS) provides compelling evidence in support of this hypothesis, revealing 3 single nucleotide polymorphisms (SNPs) in high linkage disequilibrium located within an 18 Kb segment of the T-cell receptor α locus.¹⁵ This receptor recognizes antigens presented by HLA class II molecules, such as DQB1*0602. Higher levels of surface expression of HLA DQB1*0602 on antigen-presenting cells in the homozygous state could increase the likelihood that hypocretin producing cell autoantigens are presented to these encephalitogenic T-cells over and above the hemizygous or null states. This GWAS also demonstrated a dosage effect to narcolepsy risk for SNP r1154155[C].¹⁵ Similar or unique effects of HLA-DRB1*1501, which is in tight linkage disequilibrium with DQB1*0602, may also influence the narcolepsy phenotype. Indeed, some HLA haplotypes are protective and necessary to maintain peripheral tolerance to native epitopes. To date, autoantibodies against the hypocretin system have not been discovered, but research is ongoing.¹³ Recently, evidence of a streptococcal immune response was found in newer onset cases of hypocretin deficient, cataplexy positive, HLA-DQB1*0602 positive narcolepsy suggesting that, in some cases, narcolepsy may be a post-streptococcal syndrome involving destruction of hypocretin neurons through molecular mimicry.⁴³ In that study, HLA-DQB1*0602 positive controls also demonstrated slightly higher anti-streptococcal immune titers. HLA-DQB1*0602 is known to protect against streptococcal septic shock⁴⁴ and rheumatic fever.⁴⁵ Taken together, these findings suggest that the disease-modifying aspects of HLA-DQB1*0602 may be mediated through haplotype specific effects on the streptococcal immune response. Confirmation of this hypothesis awaits further research.

The utility of HLA genotyping as a diagnostic tool in narcolepsy is controversial. HLA-DQB1*0602 is tightly associated with narcolepsy with cataplexy, present in 76% to 100% of subjects, depending on the strictness of the cataplexy definition.^46–48 The association is less robust in narcolepsy without cataplexy, present in 41% of subjects.⁴⁸ Importantly, 24% of the general population is estimated to be positive for this allele, and cataplexy-like symptoms are endorsed by up to 18% of normal subjects when experiencing intense emotions.²⁹ Therefore, HLA genotyping has been most consistently associated with narcolepsy in patients with clear-cut cataplexy, a clinical scenario in which genotyping is unnecessary to make the diagnosis. In the absence of cataplexy, the specificity of genotyping appears insufficient to distinguish narcolepsy from controls, with the majority of HLA-DQB1*0602 positive individuals under this scenario likely having another explanation for their sleepiness. In some instances cataplexy is questionable, and MSLT testing may be prohibited because of behavioral or medication-related issues. Here HLA genotyping may play a role in narcolepsy diagnosis, although CSF hypocretin-1 testing would be the test of choice in this scenario. Regardless of the diagnostic utility of HLA genotyping, the presence of HLA-DQB1*0602 likely identifies an etiologically distinct and symptomatically more severe narcolepsy phenotype.

Several limitations relating to the general design of our study are notable. Narcolepsy diagnostic precision was compromised by lack of objective diagnostic testing. We addressed this limitation in three ways. First, when available, medical record extraction confirmed the narcolepsy diagnosis. Second, we performed our analysis in a cataplexy-restricted subsample. Cataplexy is pathognomonic for narcolepsy, virtually unheard of outside the context of narcolepsy. This fact is acknowledged by the ICSD-2, which establishes sleepiness with cataplexy as adequate to diagnose narcolepsy, with confirmatory objective testing suggested but not required.³⁰ With the exception of age of onset, our findings in the general narcolepsy sample were confirmed in our cataplexy-restricted sample. Although the lack of a uniform clinical definition of cataplexy may have resulted in misspecification in some patients, we feel our cataplexy definition was robust and supported by the literature,²⁹ and that the reduced 63% HLA-DQB1*0602 positivity in our narcolepsy with cataplexy group was likely due to the population-based nature of our study. Lastly, misspecification from lack of objective testing would have included non-narcolepsy patients among our case group. We assessed sleepiness, sleep latency, and sleep duration by HLA status in a cohort of over 400 healthy controls and found no association with HLA-DQB1*0602 status (data not shown). Therefore, misspecification would likely have attenuated our results. Other limitations include the fact that our outcomes were subjective. Future studies assessing objective outcomes such as driving simulation or neuropsychological testing by HLA status in a dose response manner in a homogeneous narcolepsy sample are needed.

In conclusion, we found a dose-response relationship between HLA-DQB1*0602 allele status and measures of narcolepsy symptom severity including sleepiness, cataplexy, age of symptom onset, sleep latency, and nap and accident propensity. This suggests that HLA-DQ may function as a disease-modifying gene and raises the possibility of assessing allele status in patients for prognostic and counseling purposes. Future studies assessing if patient management strategies based on allele status is effective in improving patient safety and quality of life are needed.

DISCLOSURE STATEMENT

This was not an industry supported study. The authors have indicated no financial conflicts of interest.